Howard A. Denemark †

A number of authors have attempted to define the differences between legal and scientific methods of finding facts.¹ More than an academic exploration of two separate ways of defining truth, this comparison highlights some difficulties in using scientific evidence in court. In the words of Forensic Science Professor John Thornton:

The Court in Daubert faced the issue of whether a plaintiff, alleging harm from exposure to a prescription drug, could introduce into evidence a controversial technique of mathematical analysis in order to show causation.⁸ The Court undercut the Frye requirement that trial judges look to general acceptance of the technique by experts in the relevant field. The Court did, however, admonish judges to explore carefully the expert testimony proffered.⁹

Trial judges are left with a difficult task-as scientific, mathematical, and technical lay persons, they must analyze opinions involving matters far beyond their knowledge. When the evidence comes from scientific sources, judges will be required to understand not only the specific scientific evidence, but also the world of science, since the Supreme Court has now instructed trial judges to determine, as an admission threshold, whether the evidence is in fact "scientific knowledge."¹⁰ The Court directed judges that scientific knowledge must be learned through scientific methods, and referred to several books on the philosophy of science that are not likely to be found in courts' libraries.¹¹

A judge searching the legal literature for guidance on how scientific fact-finding differs from legal fact-finding will encounter numerous statements that science exists for the pursuit of truth, whatever its source and wherever it may lead, while law uses truth only as a stepping-stone toward its true goal of justice.¹² One particularly articulate formulation comes from Professor Milton R. Wessel:

This article will compare fact-finding in a courtroom with fact-finding in science. The scientific fact-finding focus will not be on how individual scientists conduct experiments or convince themselves of scientific facts. Rather, the inquiry will be how the institutions of science determine the state of scientific knowledge. The article will then explore the heuristic found in the legal literature and in court decisions that the difference between legal and scientific inquiry is that law seeks justice while science searches for truth. The basis of the heuristic, the myth that science is entirely factual, cold, and separate from other human endeavors, will be explored and rejected. Finally, the article will conclude that some useful comparisons and contrasts may be drawn to help those familiar with legal fact-finding identify scientific knowledge. However, the heuristic "science seeks truth; law seeks justice" should be rejected.

Many articles have contrasted the individual scientist's approach to fact-finding with the law's approach.¹⁶ These analyses have compared courtroom procedures by which parties to litigation convince the neutral trier of fact, with the way in which individual scientists themselves become convinced of scientific facts. This focus appears less useful than contrasting courtroom fact-finding with how "science," as opposed to an individual scientist, adopts certain information as fact.¹⁷ A given scientist can come to believe many things that science as an institution does not accept.¹⁸ Particularly in light of the Supreme Court's recent directive that novel scientific evidence should be admitted only if the trial judge finds it is "scientific knowledge"-which seems to imply that the trial judge should find it acceptable to science generally rather than to a given scientist who could be persuaded to act as a witness-an exploration of how "science" decides what is fact is far more valuable than an analysis of how individual scientists make that determination.

Trial judges exclude objectionable evidence so that the trier of fact will not be allowed to hear it or, if it is heard, instruct the trier to ignore that evidence.²³ In an extreme case the admission of such evidence would be grounds for a mistrial which would bring in a new trier of fact whose decisions would not be tainted by exposure to the faulty evidence.²⁴ Courts of appeal review trial judges' evidentiary rulings and juries' findings (albeit under very deferential standards of review),²⁵ acting as the final arbiters of what evidence is admissible. The courts of appeal act not only to achieve just results in the cases before them, but also to set down rules for cases yet to come.²⁶

The institution of journal science imposes at least two barriers to publication. First, given that professional reputations rest on the quality and quantity of one's publications, scientists may hesitate to submit what they consider inadequate work to peer-reviewed journals. Second, an editor may choose not to publish a piece, preventing it from coming before the community of professionals with the same authority as information published in reliable, peer-reviewed journals.³³

Publication alone cannot elevate a scientist's work to the level of an accepted scientific finding; rather it merely begins the critical process by which the transformation is possible.³⁴ Because even competent, well-trained, honest experimenters make errors in experimental procedures or in drawing deductions from their data, science journals publish much that will someday be revealed as misleading or untrue.³⁵ Only when a scientist places work in professional, peer-reviewed journals so that its limitations can be tested by other scientists, and only when it passes those tests, does the view of the scientist rise to the level of "accepted scientific fact."³⁶

The scientist, when told that the speaker's relative had a bad experience with a certain car model, would not necessarily deny the truth of the story, but would reject it as proof that other models of the car also would be defective. The result would be "proved" to the scientist's satisfaction if sufficient tests were conducted on other cars of the same model or if further data were obtained and analyzed.⁴³ Then the scientist typically would issue a statement of probability about how likely a car of this model is to have certain problems, since science tends to express its findings mathematically and probabilistically.⁴⁴

Science is far more than a collection of facts, mechanically displayed.⁵⁴ Rather, science always includes a search for order among facts.⁵⁵ A given set of scientific observations can support any number of alternative hypotheses which the scientist must eliminate from consideration.⁵⁶

In the works of Charles Darwin, for example, one finds the organization of facts far more important than the physical facts themselves. In Darwin's time there was little doubt that physical characteristics of animals had changed over time and varied with geography.⁵⁷ Fossil records showed that species exhibited change and that whole species had disappeared. A lively debate existed as to why this should be so.⁵⁸

One possible explanation was that acquired characteristics of individual animals could be passed on to their young.⁵⁹ This theory posits, for example, that animals living in cold environments will give birth to offspring with thicker fur or skin because the animal so desires. Thus, an animal's desires cause it to change or select the characteristics it passes on.⁶⁰ An Australian scientist, Paul Kammerer, attempted to prove this hypothesis in the 1920s through his work on "midwife toads."⁶¹ These toads breed on land, but a few of them are born with rough pads on their front feet which are necessary for mating in water. To prove that the needs of one generation cause genetic alterations in the next, Kammerer forced the toads to live in a water environment. He raised the eggs of the few toads that were able to breed under these difficult conditions and repeated the process for a few generations, after which a large percentage of the toads were born with the rough pads rare in midwife toads living in nature.⁶²

Kammerer thought he had proved his hypothesis-that the toads in some way understood that pads were necessary, and so actively passed on to future generations this useful anatomical change. However, Kammerer's work just as easily, and more accurately, illustrates the Darwinian model of evolution. In Darwin's terms, Kammerer imposed a strong selection pressure that allowed only those toads that carried the gene for rough pads to survive. After a few generations of mating only toads carrying the relevant gene, the population had a much higher incidence of rough pads. Understood in evolutionary terms, the experiment confirmed natural selection as the mechanism of change in species rather than some desire to pass along useful traits.⁶³

Charles Darwin fit the physical facts known to many "naturalists" of his generation into a framework that explained them in terms of natural and observable forces.⁶⁴ His pivotal work, On the Origin of Species, contains a hodgepodge of facts about pigeons, woodpeckers, plant seeds, blind cave fish, ants, bees, and the predator-prey relationship between deer and wolves.⁶⁵ Mere facts, however interesting, could not have won for Darwin's book its status as a world-changing scientific work.⁶⁶ Rather, Darwin's enduring contribution to science was the organization of diverse data into a general theory that explained how species developed. Without adding a single new fact of any consequence, On the Origin of Species became the cornerstone of modern biology.⁶⁷

The attempt to reduce science to a collection of facts, or to reduce the work of scientists to seeking physical facts, ignores the important role of theory in science. Just as history is far more than a list of dates paired with occurrences, science is broader than a gathering of facts. One could know the date of each event in the life of a country and still know little of its history.⁶⁸ Similarly, the ordering of facts into a theory reflects the goal of science far more than the discovery of the facts themselves.

The job of science does not end even after the theory is developed. Some sciences, like meteorology, then seek to predict future events.⁶⁹ Finally, some sciences seek to enable us to exert control over our world.⁷⁰ In no way, however, can one say that the ultimate purpose of science is to collect and display facts, or that the process of science stops as to a given question when enough physical facts are unearthed.

One of the most extensive explorations of the neutrality of scientists was conducted by Professor Ian Mitroff, who undertook interviews with scientists involved in lunar research just before the first American moon landing.⁷² Scientists had theorized for years on what moon rocks would reveal about the origin of the moon, our solar system, and perhaps the rest of the cosmos. The chance to record people's beliefs first in the absence of crucial data, and then again after the rocks were analyzed, was an ideal way to study how scientists change their views as new evidence develops.

Professor Mitroff asked scientists a series of questions, including one about whether scientists were objective, engaged in a disinterested search for truth. He reported:

Even if it were true that the institution of science is withdrawn from worldly concerns,⁸² modern ethics would not exempt it from moral obligations. Science is treated as part of society and remains subject to society's ethical rules. At the Nuremberg trials that followed the Second World War, some Nazi medical researchers whose experiments killed or crippled unwilling subjects argued that they were ordered to conduct certain experiments for the good of the state and in the pursuit of useful knowledge.⁸³ They claimed that judging the morality of research was not part of their roles as scientists. If society needed knowledge and their government ordered them to get it, then they had met their ethical obligations by doing the research.⁸⁴

This attempt to apply logical positivism to justify the murder of unwilling experimental subjects was rejected, and the doctors who killed their experimental subjects were convicted of crimes against humanity.⁸⁵ Those scientists and doctors indeed sought facts, but their actions were to be judged within a broader social context. As this extreme example shows, science is considered part of the larger world outside the laboratory walls.⁸⁶

Governments create laws, and then create courts to adjudicate and administer laws. Courts in America then dictate to law enforcement agencies that certain techniques of gathering evidence are impermissible. No judge or court marshal is present when law enforcement agents decide to gather evidence in an impermissible way. In the sense of direct intervention, courts cannot stop certain kinds of misconduct. Nonetheless, they purport to dictate behavior to agents in the field.

When the standards of conduct courts establish are violated, courts may respond by excluding illegally obtained evidence from consideration. It is not the underlying facts themselves that a court refuses to consider, but rather the court excludes certain facts because of the way in which they were discovered.⁹⁷ Evidence obtained illegally might be excluded, but evidence relating to the same ultimate issue would be admitted if it had been obtained in a proper manner.⁹⁸

The law, like any profession or area of specialized knowledge, defines what is and what is not reliable information, and will exclude the unreliable with little anguish.¹⁰⁴ But if the only reason to exclude coerced confessions is their unreliability, as Stein held, then a reliable coerced confession should be admitted into evidence. Even if a confession wrung from torture-battered lips is inherently untrustworthy because the victim might say anything to stop the pain, such a confession might nevertheless be reliable if it contains details known only to the perpetrator. Exclusion is more difficult to justify when a coerced confession is reliable.¹⁰⁵ The law is not removing unreliable evidence in such a case, but rather excluding valuable evidence in order to serve some other, presumably higher, goal.

The Stein case was followed for less than a decade. In 1961 the United States Supreme Court retreated from any suggestion that reliable coerced confessions might be admissible.¹⁰⁶ By 1964, the Court held explicitly that admitting involuntary confessions against a criminal defendant would not be allowed because of the "strongly felt attitude of our society that important human values are sacrificed where an agency of the government, in the course of securing a conviction, wrings a confession out of an accused against his will. . . . "¹⁰⁷ The Court went on to reiterate its belief that " 'in the end life and liberty can be as much endangered from illegal methods used to convict those thought to be criminals as from the actual criminals themselves.' "¹⁰⁸ Thus, the Court justified excluding possibly probative, reliable evidence by saying that failure to do so would result in a decline in our life, liberty, and level of civilization.

Another concern cited by courts as a primary reason for excluding involuntary confessions is that police must be deterred from violating the rights of those who are questioned in the future.¹⁰⁹ The exclusion of evidence thus acts to deter the police, whose job is to discover facts to be presented to the courts, from acting in ways the law deems undesirable.¹¹⁰

Another reason for the rule was advanced in Elkins. The Court noted that "the imperative of judicial integrity" required exclusion of unconstitutionally obtained evidence.¹¹⁷ That is, courts and agents of the government must act within the law to foster respect for laws. Like the doctrine of unclean hands, by which courts will not sully themselves by considering the arguments of litigants who have misbehaved in relation to the conduct underlying equitable claims, the Court did not wish to become a partner in the misconduct of police who deprive a defendant of the right to freedom from unreasonable searches.¹¹⁸ The Supreme Court feared that if the government refuses to obey its own laws, then disregard, disrespect, and decay must surely result.¹¹⁹

Some of the rules followed by scientific researchers interfere with their finding scientific truth. For example, the American Medical Association promulgated a code of conduct in 1946 regarding human experimentation.¹²¹ The code provided that research involving human subjects is ethical only if (1) voluntary consent of the subjects is obtained; (2) the danger of each experiment has been explored through animal experimentation; and (3) the experiment is performed under "proper medical protection and management."¹²² At least thirty-three codes of experimental ethics have been drafted since the Second World War, and informed consent is a generally accepted element of those codes.¹²³

The requirement of voluntary informed consent is a substantial barrier to human research. First, when asked whether one would like to be the subject of an experiment, one would most likely answer no. Few of us wish to be manipulated or treated as commodities, and that is precisely what many fear from agreeing to be an experimental subject. Second, genuine consent requires that the researcher give the potential subject a great deal of information and a genuine choice about participation in the study.¹²⁴ Not only does this procedure burden researchers with a time-consuming task, but it also reduces the population of those willing or able to become research subjects. The requirement of consent eliminates easy access to research subjects otherwise easiest to use-the poor who come to medical clinics but have no private doctors, prison inmates, and those who are institutionalized and unable to make their complaints heard.¹²⁵

One consequence of ethical research codes is that researchers must not test drugs on pregnant women.¹²⁶ A serious concern is that any drug given to pregnant women might harm their babies. The most direct way to find out whether a given drug harms babies in utero would be to give the drug to some pregnant women, give a placebo to other pregnant women, and monitor differences in the health of their offspring.¹²⁷ This procedure would undoubtedly lead to valuable and reliable scientific knowledge. It is, however, morally repugnant, since it places at risk innocent children whose welfare is not enhanced by this type of experiment. Scientists choose not to seek information about drugs pregnant women may use in this reprehensible way.¹²⁸ Instead they do animal studies; chemical analyses of drugs to compare drugs known to cause harm with the drug under investigation;¹²⁹ or complicated statistical analyses of public health data to determine whether a given drug can damage fetuses.¹³⁰ The ethical limitations on testing these drugs leave questions unanswered about the safety of some drugs.¹³¹ Nevertheless, scientists stay within those guidelines, sacrificing a certain amount of knowledge for higher moral values.¹³²

Codes of conduct that make scientific information gathering less efficient parallel legal rules that prevent law enforcement officers from seeking evidence in certain efficient but undesirable ways. In both cases a profession is limiting the flow of information it needs to make accurate decisions because accepting the information violates a duty higher than that represented by seeking truth. Law, in seeking justice, may deny itself the information learned from involuntary confessions and illegally seized evidence. Science denies itself the opportunity to study certain problems directly in deference to moral values. Thus, legal authors who claim that science recognizes no values higher than truth have overlooked research ethics and their codification-the scientific codes of conduct that limit researchers' truth-seeking abilities.¹³³

The twenty-two scientists raised questions about the reliability of the data and the ethics of using it.¹³⁹ Like the court in Stein, the scientists argued that the data were "almost certain to be inherently flawed" because they were obtained through unethical human experiments.¹⁴⁰ If one accepted the proposition that unethical experiments were very likely to yield unreliable data, then the question of whether to cite such experiments ceases to be a moral issue and is resolved in the relatively comfortable context of excluding unreliable evidence.¹⁴¹ The scientists' first argument was based on the morally neutral issue of reliability of data, not the ethics of citing them.

The letter further argued that using the Nazi data implicitly encourages others to perform unethical human experiments.¹⁴² Like courts seeking to discourage police from abusive practices, the scientists feared that using unethically obtained information would encourage future abuse, since the data obtained might be considered useful to the scientific community and thus bring the researchers career-enhancing recognition.¹⁴³

Another argument presented in the letter is that the use of such data debases the society using them.¹⁴⁴ This argument is indistinguishable from the courts' reasoning to exclude illegally obtained evidence because doing so would dishonor courts and debase society.¹⁴⁵

The controversy over the E.P.A. phosgene gas report demonstrates that scientists may react much as judges do in the face of information gathered unethically. Not only do they seek to "exclude" the data from the legitimate sources of information in their profession, but they justify their exclusion using the same arguments that have long been a part of legal analysis.

In the late 1970s the question of the ethics of citing the Dachau hypothermia data surfaced when a researcher at the University of Minnesota, Dr. Richard Pozos, was working on the question of how best to treat people immersed in cold water for long periods of time.¹⁴⁸ Dr. Pozos had access to the Nazi data, as did anyone interested in hypothermia, but found it troubling that the data were in use.¹⁴⁹ He consulted a bioethicist at his university, and much publicity resulted from their subsequent investigation of the problem.¹⁵⁰ Eventually a bioethics conference was held to air all sides of the dispute.¹⁵¹

One fact is clear from captured German records-the principal investigator, Sigmund Rascher, conducted both scientific and personal fraud.¹⁵² For example, Rascher faked results in a test of a supposed anti-infection drug.¹⁵³ He and his wife also denied having had a miscarriage and instead substituted another couple's child as their own so they would receive a customary gift from Himmler.¹⁵⁴

Predictably, articles written by participants at the bioethics conference disagree on whether the Nazi hypothermia data are useful¹⁵⁵ and if useful, whether they should be cited. Scientists and ethicists spoke on both sides of the dispute. Some argued that using the data to save lives would be the only fitting tribute to what would otherwise have been meaningless suffering. Others argued that trying to squeeze a profit out of the torture of innocent lives is no tribute at all.¹⁵⁶ Some argued for the use of the data, but only with a full explanation of the wrongs done to the victims.¹⁵⁷ Some stirring statements against using the data came from survivors of brutal experiments.¹⁵⁸

Perhaps no consensus was reached, but concern for these issues has continued after the conference closed.¹⁵⁹ And while it is not possible to conclude that journal science, the gatekeeper of what is and is not scientific fact, has decided to ban unethically obtained evidence, there is a high level of discomfort with publishing such information. Dr. Pozos proposed republishing the Nazi data in the New England Journal of Medicine. The editor of the Journal, one of the world's most prestigious medical and scientific journals, responded:

The editor's statement is grounded in the language of reliability, but nevertheless retains an air of moral judgment. The statement may be a convenient way of wrapping a difficult moral issue in the more secure logic of excluding unreliable evidence. The statement, "how can anyone believe it? How can anyone want to believe it?" is not the bland wording of a letter rejecting an article because its data appear insufficient.

It is most likely that if faced with proof that the data were reliable and valuable, the Journal would have shifted to the moral dimension of citing the Nazi data, much the way that the Supreme Court changed from rejecting coerced confessions because of their unreliability to rejecting them on ethical policy grounds.¹⁶⁴ Thus, as with the phosgene research example, science may ultimately deny unethically obtained evidence a place in the literature out of a desire to achieve a higher moral objective.

The research design included giving untested, untreated blood to children in the experiment's "control group." Predictably, some of these children died as a result of their exposure to virus X. The New England Journal refused to publish the report on the grounds that the researchers knowingly exposed some of their subjects to unacceptable risks.¹⁶⁹

The researchers pled that their article should have been published to disseminate information about their new, lifesaving technique. They also pointed out that, in their geographic area, blood was not routinely tested for virus X; thus the experiment benefited some children while leaving the rest no worse off than if the experiment had not been conducted.¹⁷⁰ Despite these arguments, the Journal's editors refused to publish a study that exposed some children to a risk that could have been avoided. The Journal's editors felt that the researchers' ethical duty was to change the local practice of giving untested blood to vulnerable newborns rather than to attempt to secure for themselves a publication credit in an international journal.¹⁷¹

In addition to the differences between the two disciplines stated above, the legal literature includes numerous statements implying or asserting that science is just a cold collection of facts, neither touching nor touched by the society in which it develops. This view leads to the notion that science exists in a vacuum, dispassionately seeking facts whatever the consequences, while law strives toward justice. The "law seeks justice; science seeks truth," analysis, announced by many legal authors and applied by several courts, masks some striking similarities between legal and scientific fact-finding.

Just as courts refuse to hear certain evidence deemed unreliable, so science journals, the gatekeepers of scientific information, strive to eliminate unreliable evidence. Courts block admission of illegally obtained evidence for two policy reasons: first, to discourage future misconduct by those, such as police officers, who must find facts, and second, because allowing the evidence to be used would degrade our courts and society. Science applies precisely the same logic in banning unethically obtained data from publication: the desire to deter unethical research and to preserve the moral integrity of the institutions of science underlies scientific "exclusions of evidence." One finds in both law and science the desire to avoid excluding evidence solely on moral grounds by looking initially to questions of reliability. Only when evidence cannot be shown unreliable are the necessary moral judgments made.¹⁸⁰

The law's increasing reliance on expert testimony, and the Supreme Court's recent mandate that federal judges decide what is and is not "scientific knowledge," require that the legal literature present an accurate and informed view of science and the ways in which its logic may differ from the legal analysis with which judges and lawyers are familiar. Rejecting the "law seeks justice; science seeks truth" heuristic will aid in providing the legal community with a more accurate understanding of the differences between legal and scientific fact-finding.